# AnyPhoto: Multi-Person Identity Preserving Image Generation with ID Adaptive Modulation on Location Canvas

Source: [AnyPhoto: Multi-Person Identity Preserving Image Generation with ID Adaptive Modulation on Location Canvas](<https://www.sota2.com/research/paper/69b8e0240636e5bf4afbec3f>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-18T05:00:11.457Z

Authors: Longhui Yuan

Published: 2026-03-16T03:08:38.000Z

## Abstract

Multi-person identity-preserving generation requires binding multiple reference faces to specified locations under a text prompt. Strong identity/layout conditions often trigger copy-paste shortcuts and weaken prompt-driven controllability. We present AnyPhoto, a diffusion-transformer finetuning framework with (i) a RoPE-aligned location canvas plus location-aligned token pruning for spatial grounding, (ii) AdaLN-style identity-adaptive modulation from face-recognition embeddings for persistent identity injection, and (iii) identity-isolated attention to prevent cross-identity interference. Training combines conditional flow matching with an embedding-space face similarity loss, together with reference-face replacement and location-canvas degradations to discourage shortcuts. On MultiID-Bench, AnyPhoto improves identity similarity while reducing copy-paste tendency, with gains increasing as the number of identities grows. AnyPhoto also supports prompt-driven stylization with accurate placement, showing great potential application value.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Identity-Preserving Image Generation on MultiID-Bench 2-people](<https://www.sota2.com/research/sota/identity-preserving-image-generation-on-multiid-bench-2-people.md>): Identity-preserving Image Generation; MultiID-Bench 2-people; Sim(GT): 0.401.

- [Identity-Preserving Image Generation on MultiID-Bench 3-and-4-people](<https://www.sota2.com/research/sota/identity-preserving-image-generation-on-multiid-bench-3-and-4-people.md>): Identity-preserving Image Generation; MultiID-Bench 3-and-4-people; Sim(GT): 0.424.

- [Identity-Preserving Image Generation on MultiID-Bench 1-people](<https://www.sota2.com/research/sota/identity-preserving-image-generation-on-multiid-bench-1-people.md>): Identity-preserving Image Generation; MultiID-Bench 1-people; Sim(GT): 0.448.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.14770>)

[PDF](<https://arxiv.org/pdf/2603.14770.pdf>)